Interpreting Latent Spaces of Generative Models for Medical Images using Unsupervised Methods
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Standard
Interpreting Latent Spaces of Generative Models for Medical Images using Unsupervised Methods. / Schön, Julian; Selvan, Raghavendra; Petersen, Jens.
Deep Generative Models: Second MICCAI Workshop, DGM4MICCAI 2022, Held in Conjunction with MICCAI 2022, Singapore, September 22, 2022, Proceedings. Springer, 2022. p. 24-33 (Lecture Notes in Computer Science, Vol. 13609).Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Interpreting Latent Spaces of Generative Models for Medical Images using Unsupervised Methods
AU - Schön, Julian
AU - Selvan, Raghavendra
AU - Petersen, Jens
PY - 2022
Y1 - 2022
N2 - Generative models such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs) play an increasingly important role in medical image analysis. The latent spaces of these models often show semantically meaningful directions corresponding to human-interpretable image transformations. However, until now, their exploration for medical images has been limited due to the requirement of supervised data. Several methods for unsupervised discovery of interpretable directions in GAN latent spaces have shown interesting results on natural images. This work explores the potential of applying these techniques on medical images by training a GAN and a VAE on thoracic CT scans and using an unsupervised method to discover interpretable directions in the resulting latent space. We find several directions corresponding to non-trivial image transformations, such as rotation or breast size. Furthermore, the directions show that the generative models capture 3D structure despite being presented only with 2D data. The results show that unsupervised methods to discover interpretable directions in GANs generalize to VAEs and can be applied to medical images. This opens a wide array of future work using these methods in medical image analysis.
AB - Generative models such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs) play an increasingly important role in medical image analysis. The latent spaces of these models often show semantically meaningful directions corresponding to human-interpretable image transformations. However, until now, their exploration for medical images has been limited due to the requirement of supervised data. Several methods for unsupervised discovery of interpretable directions in GAN latent spaces have shown interesting results on natural images. This work explores the potential of applying these techniques on medical images by training a GAN and a VAE on thoracic CT scans and using an unsupervised method to discover interpretable directions in the resulting latent space. We find several directions corresponding to non-trivial image transformations, such as rotation or breast size. Furthermore, the directions show that the generative models capture 3D structure despite being presented only with 2D data. The results show that unsupervised methods to discover interpretable directions in GANs generalize to VAEs and can be applied to medical images. This opens a wide array of future work using these methods in medical image analysis.
KW - eess.IV
KW - cs.CV
KW - cs.LG
U2 - 10.1007/978-3-031-18576-2_3
DO - 10.1007/978-3-031-18576-2_3
M3 - Article in proceedings
SN - 978-3-031-18575-5
T3 - Lecture Notes in Computer Science
SP - 24
EP - 33
BT - Deep Generative Models
PB - Springer
Y2 - 22 September 2022
ER -
ID: 335278278